In considering chronic interfaces, it is useful to draw distinctions between short-term interfaces lasting several days to weeks and long-term interfaces lasting months to years. In the former case, the probe’s biocompatibility is generally not the critical issue underlying effectiveness; current probe designs are sufficiently biocompatible for many applications. In the latter case, the probe’s biocompatibility is the critical issue. The working hypothesis is that microscale reactive tissue responses to the implanted device govern the effectiveness and longevity of the device. This is especially pertintent for spike recording, which has the most stringent signal transduction requirements for the neural interface. Characterizing these tissue reactions and then developing probes and strategies to optimize the performance of the electrical and/or chemical interfaces is one of the grand challenges—and opportunities—in the field.

Through a growing body of work we know that long-term neural interfaces for spike recording are feasible using the microscale probe systems that are available today. A number of investigators have recently reported viable recordings extending for months or longer using microwire arrays (Williams, J. C. 1999) and Utah silicon electrode arrays (Rousche, P.J. 1998). Our own work reports long-term spike recordings for more than six months using Michigan silicon probes in rat cerebral cortex (Vetter 2002; Kipke 2003) and, more recently, similar performance in monkeys in collaboration with Prof. Schwartz at Pittsburgh and Prof. Moran at Washington University. These recent studies are consistent with observations dating back more than 10 years (Schmidt, E.M. 1976; Schmidt, E.M. 1978). These studies can be summarized in terms of four take-home messages: (1) one can typically record useful spike activity from cerebral cortex for more than six months using any of the several types of implantable microelectrodes, (2) the recording quality is typically adequate, but not outstanding, (3) the recordings eventually degrade and ultimately fail, and (4) the unresolved issue is to increase the quality, consistency, and longevity of the recordings. These are important findings because they motivate systematic improvements to existing probe systems rather than starting from scratch using radical or unproven approaches.